A Critical Review of Theory of Mind: Are We Really Measuring What We Think We're Measuring

This study addresses the lack of consistent definitions for theory of mind (ToM). In order to better understand ToM as a construct, this study reviews previous ToM research and proposes several, theoretically-based models of different specific, general latent variables which include cognitive and affective ToM, type 1 and type 2 ToM, verbal and nonverbal ToM, and unidimensional ToM. This study tests which, if any, of these specific, general latent variables best reflects a set of diverse ToM tasks using correlated factors models, bifactor models, a general model, and an exploratory factor analysis. Results from each of these analyses consistently suggest that ToM is best reflected by a general ToM variable without specific latent variables. This suggests that it is inappropriate to characterize certain ToM measures as cognitive or affective, type 1 or type 2, or verbal or nonverbal measures. Rather, all ToM measures capture unidimensional ToM. Suggestions for future research are discussed.